Piezo Tweeter for Dogs

You miss the point.
That array (and HSS example) projects ultrasonic beam that decodes to audible sound through signal pre-procesing and air non-linearity....in effect audible sound appears as if out of thin air.
This would allow recorded audible commands to be projected, or dog disagreeable audible sounds.


You guys also are perhaps missing the point that dogs hearing goes out to 40kHz-60kHz depending on which references.
To send human inaudible sound you need purpose built US transducers and suitable amplifier....standard piezo tweeters and standard hifi amplifiers are not really fit for purpose in this application.


Perhaps a Class D amp without output filters driven by a tone or noise signal would be sufficient, economical and power efficient.
A seach on Ebay etc will reveal several types of US transducer that would likely do the job.


Dan.
 
You miss the point.
That array (and HSS example) projects ultrasonic beam that decodes to audible sound through signal pre-procesing and air non-linearity....in effect audible sound appears as if out of thin air.
This would allow recorded audible commands to be projected, or dog disagreeable audible sounds.


You guys also are perhaps missing the point that dogs hearing goes out to 40kHz-60kHz depending on which references.
To send human inaudible sound you need purpose built US transducers and suitable amplifier....standard piezo tweeters and standard hifi amplifiers are not really fit for purpose in this application.


Perhaps a Class D amp without output filters driven by a tone or noise signal would be sufficient, economical and power efficient.
A seach on Ebay etc will reveal several types of US transducer that would likely do the job.


Dan.
I wasn't sure whether you were serious or not.
I now think you may be onto something.
While searching on the internet (search efforts on and off over the last couple of years) I did come across the ultrasonic transducers (just try "ultrasonic" into google and google suggests many things, none to do with audio speaker drivers) . They can be high powered for cleaning or even welding, and I did wonder what they'd be like in air. Your suggestions are good, but how to test and measure is the problem.
 
Ultrasonic transducer can be very efficient around their peak. They're made that way.
To measure then you normally use n ultrasonic receiver.

I just googled "ultrasonic receiver", I would never have known that's what to look for.

The fence and posts are made of wood, now I'm wondering whether to go down this route instead. If the posts were made of steel it might be the best solution, but a quick idea of "burying" steel plates into the posts with transducers on top of the plates leads me to suspect that remaining wood between the plate and other side will absorb most of the energy.

Alternatively, since they're small, they would be easier to use in the small gap under the fence (gap has LoS).

Note that experimenting with an array is out of the question in my situation, the neighbour is not blind, if they see anything new anywhere on my property, they'll ask what it is. However, if someone needed distance e.g. other side of a neighbour, or if they have trees along the boundary, an array could be a good idea.

Now that I know about line transformers, i'm still thinking of piezo at this stage, but I wouldn't rule out ultrasonic transducers as a possibility.
 
There seems to be a bit of confusion about Speaker Driver power levels.
Common sense should rear it's head if an expectation of 32 W dissipation
in the APT-80 or 400 W into the Piezo is expected without quick failure.
Get out those power resistors and see what’s cooking at 32 W/ 400 W
continuous. I don’t think you are going to get the SPL’s contemplated
for this application without cooking some wieners.
 
i got 103dB out of 2.83v into one KSN1016 - wonder what a bunch could do near the same bender limit ?

I need to put these into arrays like Klipsch did

mine are Motorola

6ZMwRg3.jpg
 
There seems to be a bit of confusion about Speaker Driver power levels.
Common sense should rear it's head if an expectation of 32 W dissipation
in the APT-80 or 400 W into the Piezo is expected without quick failure.
Get out those power resistors and see what’s cooking at 32 W/ 400 W
continuous. I don’t think you are going to get the SPL’s contemplated
for this application without cooking some wieners.

The Eminence website says that specs for their speakers are from tests at rated power for 8 hours continuous. I found some similar info for the Piezo but I didn't bookmark it (quick search of KSN datasheets imply 75 Watt RMS continuous)

Either way, this application is not for continuous power, they will only be on for a few seconds at most, but IMO it's important to use the RMS Power rating to get high SPL without distortion (still not sure of what the signal content will be, several ideas in mind).
 
I'd guess a sustained 15 watts sine wave would fry an APT50 ( not sure how quick - ? - think of a 15 watt soldering pencil)

From "Power Handling" on this page of the Eminence website Understanding Loudspeaker Data | Eminence Speaker

The Eminence power rating is derived using an EIA 426A noise source and test standard. All tests are conducted for eight hours in a free-air, non-temperature controlled environment. Eminence tests samples from each of three different production runs and each sample must pass a test exceeding the rated power by 50 to 100W. The Eminence music program is double that of our standard Watts rating.


From the KSN 1165 datasheet

The 75 watt speaker incorporates protective circuitry which allows it to be used safely at continuous power levels of 400 watts.

I assume that 400W = "music power" not RMS of a sinewave or a tone
 
Right, been busy, but now that I've got some more time, I've found a good audio transformer with a datasheet (I believe that, whether datasheets tell the whole truth or not, a datasheet is better than no datasheet).

Visaton TR84
TR 84 | Visaton
http://www.visaton.de/sites/default/files/dd_product/tr84.pdf

The primary windings are
10 W / 1 kOhm;
6 W / 1,7 kOhm;
3 W / 3,3 kOhm
Secondary windings are
4; 8 Ohm
Freq response
20–20000 Hz (−3 dB)
Price in the UK about £20 GBP from several sources

So I intend to buy 1 for each KSN-1165
i.e. Amplifier output (in this case, Quad 405) goes to the 8 Ohm winding, with the 10W winding (1 kOhm) going to the KSN-1165.

The datasheet for the KSN-1165 says that its impedance appears as a 0.3 µF capacitor in series with a 30 Ohm resistor, so the question is, what (if anything) should I place in series or parallel, between or across, the TR84 and KSN-1165 ?
 
.... I've found a good audio transformer with a datasheet ....
Visaton TR84
.....
The primary windings are
10 W / 1 kOhm;
6 W / 1,7 kOhm;
3 W / 3,3 kOhm
Secondary windings are
4; 8 Ohm
......
So I intend to buy 1 for each KSN-1165
i.e. Amplifier output (in this case, Quad 405) goes to the 8 Ohm winding, with the 10W winding (1 kOhm) going to the KSN-1165.
Sorry but "too much of a good thing, becomes a bad thing"
Your turns ratio is WAY too much, over 11:1 ; you´ll damage both the tweeter (way too much drive) *and* the amplifier (way too low a load, and very reactive, just multiply capacitance by impedance ratio and divide 30 ohms by same number)

Voltage boosted piezo are often done, my former business partner wound his own air core coils for that purpose, no need of iron or even ferrite at 3kHz+ , and used about 3:1 turns ratio.

AND he put a 10 ohm 25W resistor in series with step-up air core primary, which doubled as a crossover coil, plus a series capacitor calculated again for 3kHz.

End result was an easy load for the amp , which never saw less than 10 ohms, and incredibly clean and loud treble, with wide dispersion, and remember that you are boosting Piezo voltage sensitivity by 10dB , they went from mild 92dB to scary 102dB ... compression driver territory but with extended response and better dispersion.

He had a PA rental Company and used them with the best artists who visited Buenos Aires, from Paco de Lucia to Zubin Mehta, all praising his flat, smooth, wide transparent sound, "without bullet tweeter harshness" go figure.
Not a Heavy Metal sound rig , by any means, but more suited to Jazz or Classical Music.

The datasheet for the KSN-1165 says that its impedance appears as a 0.3 µF capacitor in series with a 30 Ohm resistor, so the question is, what (if anything) should I place in series or parallel, between or across, the TR84 and KSN-1165 ?
Imagine 100 times that capacitance and 1/100th the resistance loading your amp. :(
No way to correct that, (without losing all the efficiency boost you gained) but simply do not go to such extremes.
 
Last edited:
Thanks for the info.

I'm all for winding my own, it would be cheaper for anyone in future who wants to follow the solution but is short of money (I'm not), but I don't know where to start on winding my own.

Having found out about audio line transformers, I was hoping to find something that could be used as a conventional 100v (or 70v) pair, but with a better selection of tappings on the lower impedances e.g. 4/8/16/24 Ohm . It was difficult enough finding something with a datasheet that had a spec on the freq response, and sufficient wattage.

As an alternative, could two TR84's be used, but connect with different ratios on the higher impedance windings?
e.g. amp connects to 8 Ohm on the first TR84. The 3W output of the first TR84 connects to the 10W winding on the second TR84, then the 8 Ohm of the second TR84 connects to the KSN-1165 ?
 
Forget the above.

After quite a lot of searching on the internet today, I finally found the perfect solution at AtlasIED (aka Atlas Sound).

It's their T-series of Line Matching Transformers. Very complete and comprehensive data too.

Link to the AtlasIED horns Horn Loudspeakers Designed for Applications Ranging from Paging to Sound Reinforcement and Everything in Between | AtlasIED and the high power line transformers are in the Accessories section in that link.

Link to the T-series High Power Transformers Enable Large Loudspeakers to Be Used in 70V Sound Systems | AtlasIED
Link for the datasheet for the T-series https://www.atlasied.com/f/568/ATS004129 Line Matching Transformers T Series Datasheet RevA.pdf

The 15W tapping has the lowest winding ratio in the T-series, so there is no point buying anything more than the T10 (which is the cheapest).
I took two screenshots relating to the T10 in the T-series datasheet


DtztH0E.jpg


csF442Z.jpg



The T10 is available in the US from about 25 to 40 USD, but postage to the UK is a lot of money. There is a UK distributor and I'll contact them tomorrow for pricing in the UK.

For completeness, here is the schematic that @freddi put in #104 (big thanks to freddi for the suggestion on how to connect the piezo to a domestic amp)

4brAKSe.jpg



The T10 is the transformer in the above and I'll be connecting the T10 8 Ohm to the amp, and the 14W (45 Ohm) winding to the KSN-1165.
 
There is a UK distributor and I'll contact them tomorrow for pricing in the UK.

Price in UK in GBP is slightly less (after currency conversion) than the lower USD price I found on the internet.

So I've ordered 2, but the quoted lead-time is around 6 weeks.

When I've got them, I'll start experimenting.


btw this is the spectrum of the dogs next door, about 8 metres along from the microphone, over the other side of a 6ft fence (not direct LoS).
HMYuubQ.jpg
 
Right, been busy, but now that I've got some more time, I've found a good audio transformer with a datasheet (I believe that, whether datasheets tell the whole truth or not, a datasheet is better than no datasheet).

Visaton TR84
TR 84 | Visaton
http://www.visaton.de/sites/default/files/dd_product/tr84.pdf

The primary windings are
10 W / 1 kOhm;
6 W / 1,7 kOhm;
3 W / 3,3 kOhm
Secondary windings are
4; 8 Ohm
Freq response
20–20000 Hz (−3 dB)
Price in the UK about £20 GBP from several sources

So I intend to buy 1 for each KSN-1165
i.e. Amplifier output (in this case, Quad 405) goes to the 8 Ohm winding, with the 10W winding (1 kOhm) going to the KSN-1165.

The datasheet for the KSN-1165 says that its impedance appears as a 0.3 µF capacitor in series with a 30 Ohm resistor, so the question is, what (if anything) should I place in series or parallel, between or across, the TR84 and KSN-1165 ?
There's a problem with this. 0.3uF at 20kHz is 26ohms.

You'd be better off just driving it directly from the amplifier.

It's my understanding that piezos do not have a 'power handling' as such since they theoretically use 100% reactive power and 0% real power. Instead, the limit is how much voltage you can drive them with before mechanical limits are reached or it arcs over internally and gets damaged.
 
Last edited:
Yes.
Actual "power handling" spec for classic Piezos was "35V RMS" (not a typo).

Which translates into 150W into 8 ohms, hence the nominal "150W RMS" rating, and which should be read as: "suitable for use with amplifiers rated 150W RMS into 8 ohm".

Suggested damage mechanisms are 2:

1) in theory crystal/ceramic disk flexes so much that it cracks.
Maybe so, but I never ever saw that happen.

2) mechanical flexing heats up the crystal/ceramic.
I have dissected many dead Piezos, always open, with unsoldered or broken tinsel wires.
Trying to resolder them has been tricky or useless, metal coating on disk surface is the thinnest/weakest flash possible so trying to resolder there usually means further (and unrepairable) damage.

YMMV
 
IIRC the spec for Motorola (CTS) piezos was 20 Vrms Long term and 30 Vrms for short periods. Their impedance is basically represented as a lossy capacitor but I can't rememebr whether it was around 5 nF or 50 nF but I think it is the former.
One failure mode is indeed cracking and the other one is loosing the "frozen DC bias" of the piezo ceramic due to overvoltage.

I guess that at ultrasonic frequencies the disk alone would probably work better without the horn in front.

Regards

Charles
 
Last edited: